PAOFLOW.transport.hamiltonian.operator_blc

Classes

OperatorBlockView
OperatorBlock	Container for block-structured matrix data used in quantum transport calculations.

Module Contents

class PAOFLOW.transport.hamiltonian.operator_blc.OperatorBlockView(parent, ik)

ik

parent

H

S

aux

sgm

name

dim1

dim2

class PAOFLOW.transport.hamiltonian.operator_blc.OperatorBlock(name='')

Container for block-structured matrix data used in quantum transport calculations.

This class manages related matrix data and metadata associated with a specific Hamiltonian or overlap block, including:

• Hamiltonian matrices H and overlap matrices S defined over k-points

• Optional self-energy data sgm and auxiliary workspaces aux, sgm_aux

• Dimensions, indexing information, and optional lattice vector metadata

• Flags that control which types of data are active

It supports explicit allocation and cleanup of memory for these arrays, based on the simulation’s dimensionality and configuration.

Parameters:

name (str) – Descriptive name of the operator block.

name: str = ''

dim1: int = 0

dim2: int = 0

nkpnts: int = 0

ne_sgm: int = 1

nrtot: int = 0

nrtot_sgm: int = 0

iunit_sgm: int = -1

iunit_sgm_opened: bool = False

ie: int = 0

ie_buff: int = 0

ik: int = 0

dimx_sgm: int = 0

tag: dict[str, str]

H: numpy.ndarray | None = None

S: numpy.ndarray | None = None

sgm: numpy.ndarray | None = None

aux: numpy.ndarray | None = None

sgm_aux: numpy.ndarray | None = None

icols: numpy.ndarray | None = None

irows: numpy.ndarray | None = None

icols_sgm: numpy.ndarray | None = None

irows_sgm: numpy.ndarray | None = None

ivr: numpy.ndarray | None = None

ivr_sgm: numpy.ndarray | None = None

lhave_aux = False

lhave_sgm_aux = False

lhave_ham = False

lhave_ovp = False

lhave_corr = False

ldynam_corr = False

allocated: bool = False

allocate(dim1, dim2, nkpnts, ne_sgm=1, nrtot=0, nrtot_sgm=0, lhave_aux=True, lhave_sgm_aux=False, lhave_ham=True, lhave_ovp=True, lhave_corr=False)

Allocate memory for matrix blocks and auxiliary data.

Parameters:

• dim1 (int) – Number of rows in the block.

• dim2 (int) – Number of columns in the block.

• nkpnts (int) – Number of k-points.

• ne_sgm (int) – Number of energy grid points for self-energy.

• nrtot (int) – Number of real-space lattice vectors for H and S.

• nrtot_sgm (int) – Number of real-space lattice vectors for Sigma.

• lhave_aux (bool) – Whether to allocate auxiliary matrix.

• lhave_sgm_aux (bool) – Whether to allocate auxiliary Sigma matrix.

• lhave_ham (bool) – Whether to allocate the Hamiltonian matrix H.

• lhave_ovp (bool) – Whether to allocate the overlap matrix S.

• lhave_corr (bool) – Whether to allocate correlation self-energy Sigma.

at_k(ik)

deallocate()

Deallocate all internal arrays and reset the OperatorBlock state.

This releases memory and resets allocation flags, mirroring the behavior of the Fortran operator_blc_deallocate routine.

clear()

Release all allocated memory for this block.

update(nrtot=None, nrtot_sgm=None, ie=None, ie_buff=None, ik=None, ldynam_corr=None, tag=None, name=None)

Update metadata associated with the operator block.

Parameters:

• nrtot (int, optional)

• nrtot_sgm (int, optional)

• ie (int, optional)

• ie_buff (int, optional)

• ik (int, optional)

• ldynam_corr (bool, optional)

• tag (str, optional)

• name (str, optional)

copy_from(other)

Copy all data and metadata from another OperatorBlock instance.

Parameters:

other (OperatorBlock) – The block to copy from.

copy()

memusage(memtype='all')

Estimate memory usage in MB for the specified memory type.

Parameters:

memtype ({"ham", "corr", "all"}) – Type of memory usage to compute.

Returns:

`usage_mb` – Estimated memory usage in megabytes.

Return type:

float

summary()

Return a human-readable summary of the block’s shape and configuration.

Returns:

`summary` – Multi-line string describing the operator block state.

Return type:

str